1、PDF外文：http:/ 第 1 页 附录 英文原文： A Study on AGC Scheme Based on Real Time Frequency Characteristics Hyun-ShinPark, Chungnam National University, Korea Chungnam National University, Korea kjkimcnu.ac.kr Abstract- Automatic generation control (AGC) has been one o
2、f the important issues in the operation of local area and interconnected power systems. AGC provides power demand signals for AGC power generators to control frequency. In AGC, some variables of power system such as frequency and power generation are measured to calculate the area control error (ACE
3、). The ACE value is an important factor of AGC performance. Generally, ACE value is determined by a constant frequency bias factor of power system or 1% of load method. The ACE value based on the constant frequency bias factor or1% of load method, however, may not represent the real time status of p
4、ower system correctly. The incorrect ACE value is a cause of over regulation of power system. This paper presents the method to decide actual frequency bias factor on real time frequency characteristics. In this paper, the actual frequency bias factor is determined by the measured variables such as
5、generator power output change and frequency deviation due to disturbance of power system. The actual frequency bias factor calculated is used to determine the ACE of power system. The AGC divides the calculated ACE among AGC power generators periodically. The period is determined by the time of reco
6、vered frequency. This paper also presents the simulation results for the power system model about the proposed AGC scheme. The simulation results show that the performance of AGC is improved when the proposed AGC scheme is applied to AGC. I. INTRODUCTION The scheduling and control of power ge
7、neration are vital aspects in the daily operation of power systems. Generally, energy management system is provided to control the electrical power output so as to supply ever changing system load for power as economically as possible. Automatic generation control (AGC) program 毕业设计（论文）专用纸 &n
8、bsp;第 2 页 operating within energy management system can provide the necessary means to the system operators. The AGC has been one of major issues in the operation of local control area systems and interconnected power systems for several years. Four basic objectives of power system operation d
9、uring normal operating conditions are associated with AGC: 1) matching total system generation to total system load； 2) regulating system electrical frequency error to zero； 3) distributing system generation among control areas so that net area tie flows match net area tie flow schedules
10、； 4) distributing area generation among area generation sources so that area operating costs are minimized 1. For the isolated control area, the main function of AGC is to restore the system electrical frequency to nominal frequency because there is no tie line flow among control areas 2. AGC
11、uses some field data measured through the supervisory control and data acquisition (SCADA) system which is usually implemented in energy management system. The measured data include power output of units； tie line power flow and electrical frequency of control area, etc. The power generation d
12、emand signals from the AGC function are provided to the related power generation units through SCADA system. In AGC program, the frequency of power system and interchanged power flow among control areas are used to determine area control error (ACE) value. ACE is critical for operation of AGC. In a
13、general rule, ACE calculation includes frequency deviation, power flow deviation on tieline, and prime mover control of units. AGC controller is designed to regulate ACE to zero. The frequency bias factor is an important factor for ACE calculation of control areas. The frequency bias factor is often
14、 identical to the frequency response characteristic of the control area. This paper describes the merits and demerits of the existing ACE calculation methods using constant frequency bias factor and the one percent-load method. And then this paper presents the ACE calculation method based on real ti
15、me frequency response characteristic. This paper also provides the reasonable execution period of AGC algorithm. The simulation results are also provided to demonstrate the usefulness of the proposed ACE calculation method and AGC execution period in this paper. II. EXISTING ACE CALCULATION M
16、ETHODS A. Constant frequency bias factor ACE using frequency bias factor is calculated as follows； ACE= (Pa Ps) +B (fafs) =Ptie+Bf where Pa : Actual net power interchange Ps : Scheduled power interchange fa : Measured frequency 毕业设计（论文）专用纸 第 3 页 fs : Nominal freq
17、uency B : Frequency bias factor Ptie : Power interchange deviation f :Frequency deviation Frequency bias factor, B, is determined as follows Bsys Req D Req （ R1 R2 R3 Rn） where R : Speed regulation parameter or Droop D : Load damping constant Bsys : System frequency bias factor Req : System d
18、roop Since the objective of AGC is to regulate generation of units, the input of AGC should be the actual ACE. When the frequency bias of the ACE matches with the actual frequency bias of the area, the calculated ACE and the actual ACE are identical 3. However, it is not practical to consider the fr
19、equency bias factor of (2) as the actual frequency bias, because the actual frequency bias is non-linearly related to the operating state of control area, i.e., frequency, load, generation and voltage 2. For example, the speed-droop characteristic ideally ranges from 3 to 5 percent. However, the act
20、ual speed-droop characteristic may thus exhibit incremental regulation ranging from 2 to 12 percent, depending on the unit output 2. The speed-droop characteristic also depends on the type of units, such as steam unit and hydraulic unit. For example, steam turbines have a number of control valves, e
21、ach having nonlinear flow area versus position characteristic 2. The generator response is not always predictable and has time delay. Load damping characteristic, D, is the relationship between the changes in actual load due to a change in frequency. The relationship cannot be measured and varies co
22、ntinuously with the system status. Typical values of D are 1 to 2 percent 2. Therefore, the fixed estimated frequency bias value such as (2) is not the same as the actual frequency bias. If the estimated frequency bias is much higher than the actual frequency bias, then variations in the calculated
23、ACE can be much larger than variations in the actual ACE. It causes often over-regulation. The AGC depends on the calculated ACE using constant frequency bias using system frequency, generator regulation and load damping as inputs. Therefore, it is critical to get actual frequency bias value for bet
24、ter AGC performance. In conclusion, ACE calculation method using constant frequency bias cannot represent the actual power system status correctly. B. One-percent of load method North America Electric Reliability Council (NERC) offers several possible ways to calculate frequency bias. The general practice is to establish frequency bias in each control area once a year based on the areas natural regulation characteristic (1/R+D) corresponding to the forecasted peak load of the coming year 4. It is called as “1%
